OpenAI compatible ProvidersBaseten

Baseten Provider

Baseten is a platform for running and testing LLMs. It allows you to deploy models that are OpenAI API compatible that you can use with the AI SDK.

Setup

The Baseten provider is available via the @ai-sdk/openai module as it is compatible with the OpenAI API. You can install it with

pnpm

npm

yarn

pnpm add @ai-sdk/openai

Provider Instance

To use Baseten, you can create a custom provider instance with the createOpenAI function from @ai-sdk/openai:

import { createOpenAI } from '@ai-sdk/openai';

const BASETEN_MODEL_ID = '<deployment-id>';
const BASETEN_DEPLOYMENT_ID = null;

// see https://docs.baseten.co/api-reference/openai for more information
const basetenExtraPayload = {
  model_id: BASETEN_MODEL_ID,
  deployment_id: BASETEN_DEPLOYMENT_ID,
};

const baseten = createOpenAI({
  name: 'baseten',
  apiKey: process.env.BASETEN_API_KEY ?? '',
  baseURL: 'https://bridge.baseten.co/v1/direct',
  fetch: async (url, request) => {
    const bodyWithBasetenPayload = JSON.stringify({
      ...JSON.parse(request.body),
      baseten: basetenExtraPayload,
    });
    return await fetch(url, { ...request, body: bodyWithBasetenPayload });
  },
});

Be sure to have your BASETEN_API_KEY set in your environment and the model deployment id ready. The deployment_id will be given after you have deployed the model on Baseten.

Language Models

You can create Baseten models using a provider instance. The first argument is the served model name, e.g. ultravox.

const model = baseten('ultravox');

Example

You can use Baseten language models to generate text with the generateText function:

import { createOpenAI } from '@ai-sdk/openai';
import { generateText } from 'ai';

const BASETEN_MODEL_ID = '<deployment-id>';
const BASETEN_DEPLOYMENT_ID = null;

// see https://docs.baseten.co/api-reference/openai for more information
const basetenExtraPayload = {
  model_id: BASETEN_MODEL_ID,
  deployment_id: BASETEN_DEPLOYMENT_ID,
};

const baseten = createOpenAI({
  name: 'baseten',
  apiKey: process.env.BASETEN_API_KEY ?? '',
  baseURL: 'https://bridge.baseten.co/v1/direct',
  fetch: async (url, request) => {
    const bodyWithBasetenPayload = JSON.stringify({
      ...JSON.parse(request.body),
      baseten: basetenExtraPayload,
    });
    return await fetch(url, { ...request, body: bodyWithBasetenPayload });
  },
});

const { text } = await generateText({
  model: baseten('ultravox'),
  prompt: 'Tell me about yourself in one sentence',
});

console.log(text);

Baseten language models are also able to generate text in a streaming fashion with the streamText function:

import { createOpenAI } from '@ai-sdk/openai';
import { streamText } from 'ai';

const BASETEN_MODEL_ID = '<deployment-id>';
const BASETEN_DEPLOYMENT_ID = null;

// see https://docs.baseten.co/api-reference/openai for more information
const basetenExtraPayload = {
  model_id: BASETEN_MODEL_ID,
  deployment_id: BASETEN_DEPLOYMENT_ID,
};

const baseten = createOpenAI({
  name: 'baseten',
  apiKey: process.env.BASETEN_API_KEY ?? '',
  baseURL: 'https://bridge.baseten.co/v1/direct',
  fetch: async (url, request) => {
    const bodyWithBasetenPayload = JSON.stringify({
      ...JSON.parse(request.body),
      baseten: basetenExtraPayload,
    });
    return await fetch(url, { ...request, body: bodyWithBasetenPayload });
  },
});

const result = streamText({
  model: baseten('ultravox'),
  prompt: 'Tell me about yourself in one sentence',
});

for await (const message of result.textStream) {
  console.log(message);
}

Baseten language models can also be used in the generateObject, and streamObject functions.